Gesture recognition using Kinect sensor &; M-SVM and its realization in remotely controlling a robot

• Main Authors: Chien-Hung Lai, 賴建宏
• Other Authors: Chaoming Hsu
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/34629811118299560274

碩士 === 國立嘉義大學 === 資訊工程學系研究所 === 101 === This paper uses machine learning and Kinect sensor to design a simple, convenient, yet effective gesture recognition method and its realization for a robot remote control system. The Kinect sensor is first used to capture the human body skeleton with depth information. A gesture training and identification method is designed using multiple-classes support vector machine (M-SVM) and its realized to remotely control a mobile robot for certain actions via the Bluetooth. Experimental results show that the designed method can achieve, on an average, more than 97% of accurate identification of 7 types of gestures and it can effectively control a real e-puck robot for the designed gesture commands.